Does quantum computing already possess the level of abstraction to be explicable even without knowledge of physics?

Question

Currently, quantum computer science (in contrast to classical computer science) can mostly only be understood if one has a good inside knowledge of physics, or more precisely quantum physics. Only then one can really understand the explanations of superposition, coherence, entanglement, etc. in relation to Qubits. I felt the same way and had to take the extra physics course, in order to really get into Quantum Computing.

However, I heard more and more that it is "theoretically" possible to abstract quantum computer science to such a degree, that knowledge in physics is no longer required. Is there any truth to this or is it just way too conceptual so far?

(And yes, I am aware that this question might be more related to didactics rather than on-topic with this board. But I wasn't quite sure if I should've posted it on StackOverflow. So, forgive my error, given I've done one)

The book Q is for Quantum attempts to explain quantum computing using high school math, but without watering down the main ideas. — littleO, Feb 04 '20 at 05:12
Does this answer your question? Is quantum computing mature enough for a computer scientist with no physics background? — Jonathan Trousdale, Feb 04 '20 at 17:13
I agree with Scott Aaronson when he says that quantum mechanics has been done a disservice by being exclusively taught as a field of physics, with all the baggage of atoms, Hamiltonians, and such. All that's required, IMO, to understand QC is a strong background in linear algebra, in addition to the postulates of QM (which is basically: 1. Hilbert spaces, 2. Unitary transformations, 3. Measurement operators, and 4. Tensor products of composite systems). Chapter 2 of Neilsen and Chuang covers all that. (And I don't think it says the word "electron" even once.) — Bridgeburners, Feb 04 '20 at 19:07

score 12 · Accepted Answer · answered Feb 03 '20 at 19:17

I don't think you need to know quantum physics to understand quantum computing - similarly to how you don't think about the hardware implementation of the classical computers when you write high-level code for them.

The field of quantum computing has grown to the point where one cannot really teach all of it in one course, so different approaches to teaching emerge. Some courses indeed start with quantum mechanics, introducing qubits as particles and operations as physical processes on them. Others choose to abstract the physics away and to discuss the quantum computing algorithms in mathematical terms - representing states as vectors, operators as matrices and so on. Examples include:

"Introduction to Quantum Information Science" course by Scott Aaronson
"Introduction to Quantum Information Processing" by Richard Cleve
"Quantum Computation" course by Umesh Vazirani (which mentions physics now and then but more as an illustration)
Our group took a similar approach when we taught a class "Introduction to Quantum Computing and Quantum Programming in Q#" last year.

There is plenty of material to cover even without diving deep into physics, and there are lots of programming languages that allow the learner to focus on studying and implementing the high-level algorithms instead.

I agree with this, though I've found you do pick up bits and pieces of the physics as you progress. Computation is fundamentally about information; in that respect, one can ask: Do we need to know about Landauer's principle and thermodynamics to explain how classical computers work? The answer is clearly no, you can understand gates, algorithms, and software without any purely physical concepts. I like Aaronson's book Quantum Computing Since Democritus. It's about as fun as a 'textbook' can get and focuses on quantum information with very little physics sprinkled in. — Greenstick, Feb 03 '20 at 19:41

score 5 · Answer 2 · answered Feb 03 '20 at 17:46

I believe it is possible to study Quantum Mechanics by studying Quantum Computing.

A qubit is a simplest quantum system showing non-classical behavior (superposition of basis states). It is very logical to start studying Quantum Mechanics from the simplest quantum system, and then move to more complex multiqubit systems. If you need Quantum Mechanics to understand Quantum Information Science, study Quantum Mechanics of qubits!

I like very much introductory online course on Quantum Mechanics and Quantum Computation by Prof. Vazirani, now archived, which is an interesting attempt to teach both Quantum Mechanics and Quantum Computation in a single course.

Does quantum computing already possess the level of abstraction to be explicable even without knowledge of physics?

2 Answers2